An embodiment provides methods and arrangements for visualizing and exploring natural-language text. In an embodiment, natural language text is received, and this is parsed to form a directed graph comprising a plurality of nodes. The directed graph is converted to an outline graph comprising core nodes and layer nodes. The outline graph is simplified, and there is created, for display on a user interface, an interactive visual representation of the simplified outline graph. Other variants and embodiments are broadly contemplated herein.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A method of processing and visualizing natural language text, said method comprising: utilizing at least one processor to execute computer code that performs the steps of: receiving natural language text; parsing the received natural language text to form a data forest comprising a directed graph comprising a plurality of nodes, each of the plurality of nodes being associated with (i) an identifier, (ii) a label, (iii) a value, (iv) a set of attributes, and (v) a mapping from attributes to values; converting the directed graph to an outline graph comprising core nodes and layer nodes, wherein the converting comprises (i) grouping nodes of the data forest having the same label and assigning a set of identifiers to the grouped nodes and (ii) mapping each attribute to a multiset of values for nodes of the data forest having the same attribute; and creating, for display on a user interface, an interactive visual representation of the outline graph, wherein the interactive visual representation comprises an outline graph view and an analytics tool panel, wherein the outline graph view comprises a visual representation of a summarized view of the natural language text based on the outline graph and wherein the analytics tool panel provides controls for interacting with the outline graph view; the creating comprising simplifying the outline graph by reducing the outline graph by weighting the nodes within the outline graph based upon a relative importance of the node with respect to other nodes and removing nodes having a weight below a predetermined value.
2. The method according to claim 1 , wherein the directed graph is formed based on an input schema.
3. The method according to claim 1 , wherein said parsing comprises semantic parsing.
4. The method according to claim 1 , wherein: wherein the set of identifiers and the multiset of values comprises singleton sets; and wherein said converting comprises adding edges from core nodes to layer nodes.
5. The method according to claim 1 , comprising simplifying the outline graph, wherein said simplifying comprises contracting the outline graph via combining two or more nodes sharing at least one similar characteristic.
6. The method according to claim 5 , wherein said reducing is performed on the contracted outline graph via.
7. The method according to claim 6 , wherein: said removing is governed by one or more characteristics relating to a size of the interactive visual representation to be displayed on the user interface.
8. The method according to claim 5 , wherein said contracting comprises combining two nodes given at least one of the following conditions (i), (ii) and (iii): (i) both of the two nodes are core nodes; (ii) both of the two nodes are roots in a data forest induced by the layer nodes; and (iii) both of the two nodes are sibling nodes.
9. The method according to claim 1 , wherein said creating comprises: automatically generating an extraction program based on the outline graph; and executing the automatically generated extraction program.
10. The method according to claim 1 , wherein said creating comprises permitting one or more user-initiated operations with respect to the interactive visual representation, and propagating the one or more user-initiated operations back to the directed graph.
11. The method according to claim 10 , wherein the one or more user-initiated operations comprise deleting a node in the interactive visual representation.
12. The method according to claim 10 , wherein the one or more user-initiated operations comprise defining a filter over one or more core nodes represented in the interactive visual representation.
13. The method according to claim 10 , wherein said creating comprises creating a revised interactive visual representation in response to the one or more user-initiated operations.
14. The method according to claim 1 , wherein the interactive visual representation comprises a stylized view of the outline graph.
15. The method according to claim 14 , wherein the stylized view of the outline graph comprises: a stylized representation of the core nodes in a visually central position; and a stylized representation of the layer nodes in at least one position removed from the visually central position.
16. The method according to claim 14 , wherein the interactive visual representation comprises a panel of auxiliary controls to assist a user in exploring and interacting with the stylized view of the outline graph.
17. An apparatus for processing and visualizing natural language text, said apparatus comprising: at least one processor; and a computer readable storage medium having computer readable program code embodied therewith and executable by the at least one processor, the computer readable program code comprising: computer readable program code configured to receive natural language text; computer readable program code configured to parse the received natural language text to form a data forest comprising a directed graph comprising a plurality of nodes, each of the plurality of nodes being associated with (i) an identifier, (ii) a label, (iii) a value, (iv) a set of attributes, and (v) a mapping from attributes to values; computer readable program code configured to convert the directed graph to an outline graph comprising core nodes and layer nodes, wherein the converting comprises (i) grouping nodes of the data forest having the same label and assigning a set of identifiers to the grouped nodes and (ii) mapping each attribute to a multiset of values for nodes of the data forest having the same attribute; and computer readable program code configured to create, for display on a user interface, an interactive visual representation of the outline graph, wherein the interactive visual representation comprises an outline graph view and an analytics tool panel, wherein the outline graph view comprises a visual representation of a summarized view of the natural language text based on the outline graph and wherein the analytics tool panel provides controls for interacting with the outline graph view; the creating comprising simplifying the outline graph by reducing the outline graph by weighting the nodes within the outline graph based upon a relative importance of the node with respect to other nodes and removing nodes having a weight below a predetermined value.
18. A computer program product for processing and visualizing natural language text, said computer program product comprising: a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising: computer readable program code configured to receive natural language text; computer readable program code configured to parse the received natural language text to form a data forest comprising a directed graph comprising a plurality of nodes, each of the plurality of nodes being associated with (i) an identifier, (ii) a label, (iii) a value, (iv) a set of attributes, and (v) a mapping from attributes to values; computer readable program code configured to convert the directed graph to an outline graph comprising core nodes and layer nodes, wherein the converting comprises (i) grouping nodes of the data forest having the same label and assigning a set of identifiers to the grouped nodes and (ii) mapping each attribute to a multiset of values for nodes of the data forest having the same attribute; and computer readable program code configured to create, for display on a user interface, an interactive visual representation of the outline graph, wherein the interactive visual representation comprises an outline graph view and an analytics tool panel, wherein the outline graph view comprises a visual representation of a summarized view of the natural language text based on the outline graph and wherein the analytics tool panel provides controls for interacting with the outline graph view; the creating comprising simplifying the outline graph by reducing the outline graph by weighting the nodes within the outline graph based upon a relative importance of the node with respect to other nodes and removing nodes having a weight below a predetermined value.
19. A method comprising: utilizing at least one processor to execute computer code that performs the steps of: receiving natural language text; performing semantic parsing on the received natural language text to form, based on an input schema, a data forest comprising a directed graph comprising a plurality of nodes, each of the plurality of nodes being associated with (i) an identifier, (ii) a label, (iii) a value, (iv) a set of attributes, and (v) a mapping from attributes to values; converting the directed graph to an outline graph comprising core nodes and layer nodes, wherein the converting comprises (i) grouping nodes of the data forest having the same label and assigning a set of identifiers to the grouped nodes and (ii) mapping each attribute to a multiset of values for nodes of the data forest having the same attribute; contracting the outline graph via combining two or more nodes sharing at least one similar characteristic; creating, for display on a user interface, an interactive visual representation of the outline graph, wherein the interactive visual representation comprises an outline graph view and an analytics tool panel; wherein the interactive visual representation affords development of an information extractor and comprises a stylized view of the outline graph comprising a visual representation of a summarized view of the natural language text based on the outline graph; said creating comprising permitting one or more user-initiated operations, using the analytics tool panel, with respect to the interactive visual representation, and propagating the one or more user-initiated operations back to the directed graph, wherein the creating comprises simplifying the outline graph by reducing the outline graph by weighting the nodes within the outline graph based upon a relative importance of the node with respect to other nodes and removing nodes having a weight below a predetermined value; and automatically generating an extraction program based on the outline graph.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
December 30, 2015
September 18, 2018
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.